Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FoodPuzzle: Developing Large Language Model Agents as Flavor Scientists

Tenghao Huang, Donghee Lee, John Sweeney, Jiatong Shi, Emily Steliotes, Matthew Lange, Jonathan May, Muhao Chen

TL;DR

A novel Scientific Agent approach is proposed, integrating in-context learning and retrieval augmented techniques to generate grounded hypotheses in the domain of food science, demonstrating its potential to transform flavor development practices.

Abstract

Flavor development in the food industry is increasingly challenged by the need for rapid innovation and precise flavor profile creation. Traditional flavor research methods typically rely on iterative, subjective testing, which lacks the efficiency and scalability required for modern demands. This paper presents three contributions to address the challenges. Firstly, we define a new problem domain for scientific agents in flavor science, conceptualized as the generation of hypotheses for flavor profile sourcing and understanding. To facilitate research in this area, we introduce the FoodPuzzle, a challenging benchmark consisting of 978 food items and 1,766 flavor molecules profiles. We propose a novel Scientific Agent approach, integrating in-context learning and retrieval augmented techniques to generate grounded hypotheses in the domain of food science. Experimental results indicate that our model significantly surpasses traditional methods in flavor profile prediction tasks, demonstrating its potential to transform flavor development practices.

FoodPuzzle: Developing Large Language Model Agents as Flavor Scientists

TL;DR

A novel Scientific Agent approach is proposed, integrating in-context learning and retrieval augmented techniques to generate grounded hypotheses in the domain of food science, demonstrating its potential to transform flavor development practices.

Abstract

Flavor development in the food industry is increasingly challenged by the need for rapid innovation and precise flavor profile creation. Traditional flavor research methods typically rely on iterative, subjective testing, which lacks the efficiency and scalability required for modern demands. This paper presents three contributions to address the challenges. Firstly, we define a new problem domain for scientific agents in flavor science, conceptualized as the generation of hypotheses for flavor profile sourcing and understanding. To facilitate research in this area, we introduce the FoodPuzzle, a challenging benchmark consisting of 978 food items and 1,766 flavor molecules profiles. We propose a novel Scientific Agent approach, integrating in-context learning and retrieval augmented techniques to generate grounded hypotheses in the domain of food science. Experimental results indicate that our model significantly surpasses traditional methods in flavor profile prediction tasks, demonstrating its potential to transform flavor development practices.
Paper Structure (14 sections, 3 equations, 5 figures, 2 tables)

This paper contains 14 sections, 3 equations, 5 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Task and Data
  4. Task Definition
  5. Constructing the FoodPuzzle Dataset
  6. Evaluation Protocols
  7. Evaluated Methods
  8. Baseline Methods
  9. Architecture of the Scientific Agent
  10. Experiments
  11. Main Results
  12. Error Analysis
  13. Future Opportunities
  14. Conclusion

Figures (5)

  • Figure 1: Flavor is determined by diverse flavor molecules. Sourcing and identifying these molecules from various foods is a time-consuming and resource-intensive task for food scientists. Understanding these connections is crucial in flavor science for developing and enhancing food products to ensure appealing taste experiences for consumers. In this work, we explore how LLMs can assist in this process.
  • Figure 2: A high level overview of the FoodPuzzle data hierarchy
  • Figure 3: Distribution of the number of flavor molecules in the FOODPUZZLE dataset.
  • Figure 4: PCA visualization illustrating clustering of food entities based on molecular profiles.
  • Figure 5: Architecture of the proposed Scientific Agent